Everyone Has AI Now. That's Exactly Why It Won't Win Your Tenancy Dispute

There was a moment, not long ago, when showing up to the Residential Tenancy Branch with a polished, AI-drafted demand letter might have felt like an edge. A tenant or landlord could feed a few prompts into a chatbot and walk away with something that looked like competent legal work - citations, confident language, the whole package.

That moment is over. Everyone has access to AI now. Landlords have it. Tenants have it. Property managers have it. So do the arbitrators. When everyone is using the same free tools, the tool itself stops being an advantage. What's left is the thing that always mattered: whether the underlying legal reasoning is actually sound.

And here's the uncomfortable part - AI is often very good at making bad reasoning sound excellent.

AI Is Built to Agree With You

The core problem isn't that AI is stupid. It's that AI is agreeable. These tools are designed to find a solution, reach an agreement, and give you an answer. They don't readily fight back or play the role of bad-news messenger.

That matters enormously in a legal context, because of how people actually use these tools. When someone types a prompt, the prompt is almost always loaded with the conclusion they're hoping to reach - whether they realize it or not. The answer that comes back is quietly shaped to confirm that bias.

Consider a tenant who types the following into an AI chatbot:

"My landlord is lying to me. They say their hydro bill is $200 but they refuse to send me a screenshot of the bill. I think they're pulling a fast one. Can you find me the law on this and write a demand letter?"

That prompt creates the very thing it's looking for. It walks in assuming the landlord is a liar, and the AI obligingly weaponizes the law in a direction it was never designed to go - at least in this example. The output will read as substantiated. It will sound backed-up. But at its core it may be hollow - and sometimes worse than hollow, because it sends the tenant marching toward a confrontation built on a foundation of sand.

The Output Is Not a Supreme Authority

The biggest mistake we see is people treating whatever the AI produces as the truth, as if a chatbot had handed down competent legal advice, and that the resulting threats and demands carry real authority.

They oftentimes don't. When that AI-generated product reaches the Residential Tenancy Branch, or lands on our desk as opposing lawyers, it tends to seem valid mostly because it sounds nice. For anyone with real experience in tenancy law, it doesn't take much digging to see these for what they often are: hollow threats, unsupported demands, and claims that simply can't be substantiated.

The danger is that the person relying on the output can't see that. They've effectively bet their tenancy on a long conversation with a tool whose ancestor was autocorrect, and they're hanging an entire case on it. Sometimes it works out. Often it doesn't.

How We Actually Use AI

None of this is to say AI has no place in preparing for your tenancy dispute. At Bright Law, we use it constantly to work efficiently. It's genuinely good at sifting through large volumes of material quickly and cobbling together a starting point. In fact, AI was used to transcribe - but not reason - the very information contained in this blog post.

But there's one non-negotiable rule at Bright Law: everything AI touches gets double-checked by a human who understands the law. While we do not use AI tools to assist us with our legal reasoning, we use its assistance in summarizing large volumes of information and data. Anything that it does produce is double-checked against the information it was fed, to ensure that the output is not some kind of hallucination or a misrepresentation of the information (or its conclusions).

That double-checking lens is precisely what most landlords and tenants don't have. Without the experience to spot the nuances, they can't tell the difference between a sound argument and a confident fabrication. And by the time that difference matters - at a hearing - it's too late to learn it.

The Bottom Line

You can put the entire RTB Policy Guidelines into an AI engine as PDFs. You can load in the Rules of Procedure. You can drop in whole cases. The tool will still give you back a conclusion that sounds convincing, because that's what it's built to do.

What it can't give you is judgment. Both landlords and tenants routinely jeopardize their investments by treating an agreeable chatbot as a substitute for legal advice. Use AI to get oriented if you like, but before you act on anything it tells you, get advice from someone who can tell you when it's wrong.

If you're a landlord or tenant facing a dispute and you want to know whether your position actually holds up, that's exactly what we do. Contact Bright Law before you commit to a position you can't walk back.

--

Related reading

Part of our series on the things landlords and tenants most often get wrong before they call us: